Technically not since ~2000. Windows uses UTF-16 throughout the OS. That is, Windows strings will be encoded and decoded as UTF-16 not UCS-2.
However the tricky bit is that the kernel doesn't enforce this so it's possible for programmers to intentionally make broken UTF-16 strings. A broken UTF-16 string shouldn't be considered UCS-2 just because it happens to be a valid UCS-2 string (otherwise all bit patterns could be called "UCS-2" so long as they are an even number of bytes in length).
However the tricky bit is that the kernel doesn't enforce this so it's possible for programmers to intentionally make broken UTF-16 strings. A broken UTF-16 string shouldn't be considered UCS-2 just because it happens to be a valid UCS-2 string (otherwise all bit patterns could be called "UCS-2" so long as they are an even number of bytes in length).